University of New Hampshire
University of New Hampshire Scholars' Repository
Affiliate Scholarship
Center for Coastal and Ocean Mapping
1992
SeaBeam and seismic reflection surveys on the
Ontong Java Plateau
Larry A. Mayer
University of New Hampshire
Tom H. Shipley
University of Texas at Austin
Edward L. Winterer
University of California - San Diego
David Mosher
Geological Survey of Canada
Rick A. Hagen
University of Hawaii - West Oahu
Follow this and additional works at: http://scholars.unh.edu/ccom_affil
Part of the Geology Commons, Oceanography and Atmospheric Sciences and Meteorology
Commons, and the Stratigraphy Commons
Recommended Citation
Mayer, L.A., Shipley, T., Winterer, E.L., Mosher, D., and Hagen, R., 1991, SeaBeam and seismic reflection surveys on the Ontong Java
Plateau, in Berger, W.H. and Kroenke, L., eds., Proceedings of the Ocean Drilling Program, vol. 130, pp. 45-76.
This Conference Proceeding is brought to you for free and open access by the Center for Coastal and Ocean Mapping at University of New Hampshire
Scholars' Repository. It has been accepted for inclusion in Affiliate Scholarship by an authorized administrator of University of New Hampshire
Scholars' Repository. For more information, please contact [email protected].
Berger, W.H., Kroenke, L.W., Mayer, L.A., et al, 1993
Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 130
3. SEISMIC STRATIGRAPHY OF THE ONTONG JAVA PLATEAU1
David C. Mosher,2 Larry A. Mayer,2,3 Tom H. Shipley,4 Edward L. Winterer,5 Rick A. Hagen,6 Janice C. Marsters,6
Franck Bassinot,7 Roy H. Wilkens,6 and Mitchell Lyle8
ABSTRACT
The Ontong Java Plateau, a large, deep-water carbonate plateau in the western equatorial Pacific, is an ideal location for
studying responses of carbonate sedimentation to the effects of changing paleoceanographic conditions. These carbonate
responses are often reflected in the physical properties of the sediment, which in turn control the appearance of seismic reflection
profiles. Seismic stratigraphy analyses, correlating eight reflector horizons to each drill site, have been conducted in an attempt
to map stratigraphic data. Accurate correlation of seismic stratigraphic data to drilling results requires conversion of traveltime
to depth in meters. Synthetic seismogram models, using shipboard physical properties data, have been generated in an attempt to
provide this correlation.
Physical properties, including laboratory-measured and well-log data, were collected from sites drilled during Deep Sea
Drilling Project Legs 30 and 89, and Ocean Drilling Program Leg 130, on the top and flank of the Ontong Java Plateau.
Laboratory-measured density is corrected to in-situ conditions by accounting for porosity rebound resulting from removal of the
sediment from its overburden. The correction of laboratory-measured compressional velocity to in situ appears to be largely a
function of increases in elastic moduli (especially shear rigidity) with depth of burial, more than a function of changes in
temperature, pressure, or density (porosity rebound). Well-log velocity and density data for the ooze intervals were found to be
greatly affected by drilling disturbance; hence, they were disregarded and replaced by lab data for these intervals.
Velocity and density data were used to produce synthetic seismograms. Correlation of seismic reflection data with synthetic
data, and hence with depth below seafloor, at each drill site shows that a single velocity-depth function exists for sediments on
the top and flank of the Ontong Java Plateau. A polynomial fit of this function provides an equation for domain conversion:
Depth (mbsf) = 44.49 + 0.800(traveltime[ms]) + 3.308 × 10"4 (traveltime[ms]2)
Traveltime (ms) = -35.18 + 1.118(depth[mbsf]) - 1.969 × KT* (depth[mbsf]2)
Seismic reflection profiles down the flank of the plateau undergo three significant changes: (1) a drastic thinning of the
sediment column with depth, (2) changes in the echo-character of the profile (development of seismic facies), and (3) loss of
continuous, coherent reflections. Sediments on the plateau top were largely deposited by pelagic processes, with little significant
postdepositional or syndepositional modification. Sediments on the flank of the plateau are also pelagic, but they have been
modified by faulting, erosion, and mass movement. These processes result in disrupted and incoherent reflectors, development
of seismic facies, and redistribution of sediment on the flank of the plateau.
Seismic stratigraphic analyses have shown that the sediment section decreases in thickness by as much as 65% between water
depths of 2000 m water depth (at the top of the plateau) and 4000 m (near the base of the plateau). Thinning is attributed to
increasing carbonate dissolution with depth. If this assumption is correct, then changes in the relative thicknesses of seismostratigraphic units at each drill site are indicative of changes in the position of the lysocline and the dissolution gradient between the
lysocline and the carbonate compensation depth. We think that a shallow lysocline in the early Miocene caused sediment thinning.
A deepening of the lysocline in the late-early Miocene caused relative thickening at each site. Within the middle Miocene, a sharp
rise in lysoclinal depth occurs, concurrent with a steepening of the dissolution gradient. These events result in sediment thinning
at all four sites. The thicker sections in the late Miocene likely correspond to a deepening of the lysocline, and a subsequent rise
in the lysocline again hinders accumulation of sediment in the very late Miocene and Pliocene.
INTRODUCTION
Pelagic carbonate sediments are excellent indicators of past oceanographic conditions and processes. Responses to varying ocean-
1
Berger, W.H., Kroenke, L.W., Mayer, L.A., et al., 1993. Proc. ODP, Sci. Results,
130: College Station, TX (Ocean Drilling Program).
2
Department of Oceanography, Dalhousie University, Halifax, Nova Scotia B3H 4J1,
Canada.
3
Ocean Mapping Group, Department of Surveying Engineering, University of New
Brunswick, P.O. Box 4400, Fredericton, New Brunswick E3B 5A3, Canada.
4
Institute for Geophysics, University of Texas at Austin, 8701 Mopac Boulevard,
Austin, TX 79759, U.S.A.
5
Geological Research Division, Scripps Institution of Oceanography, University of
California, San Diego, La Jolla, CA 92093, U.S.A.
6
Department of Geology and Geophysics, School of Ocean and Earth Science and
Technology, University of Hawaii, 2525 Correa Road, Honolulu, HI 96822, U.S.A.
7
Laboratoire de Géologie du Quaternaire, Case 907, CNRS-Luminy, 13288 Marseille
Cedex 9, France.
8
Borehole Research Group, Lamont-Doherty Geological Observatory, Palisades, NY
10963, U.S.A.
ographic conditions can result in changes in sedimentation regimes
and in physical properties of the sediment. Seismic reflection techniques can be used to investigate these changes remotely because
sedimentation styles are often expressed in seismic reflection patterns, and because the acoustic response of a sedimentary sequence
to an impulse is a function of physical properties changes within that
sequence. This paper attempts to accurately correlate seismic reflection data to core data based on physical properties information from
sediments of the Ontong Java Plateau collected during Ocean Drilling
Program (ODP) Leg 130. These correlations will allow interpretation
of the seismic reflection data with respect to stratigraphy, sedimentology, and paleoceanography of the Ontong Java Plateau, based on
the geology in Leg 130 cores.
Location and Geologic Setting
The Ontong Java Plateau, at 1.5 million km2 in area, is the largest
of the world's classic "oceanic" plateaus. It is located in the western
equatorial Pacific north of the Solomon Islands (Fig. 1). At the top of
33
D.C. MOSHER ET AL.
4°N
2°N
Ontong Java
Plateau
LEGEND
—
DSDP Sites 289 and 586
_
•
155 ° E
T.Washington
seismic data
Joides Resolution
seismic data
Drilling site
165° E
160° E
Figure 1. Location and bathymetric map of the Ontong Java Plateau, illustrating available seismic reflection data and the ODP Leg 130 and DSDP
drilling site locations.
the plateau, approximately 1.5 km of well-stratified Mesozoic and
Cenozoic pelagic sediments overlie basaltic crust (Shipboard Scientific Party, 1971, 1975; Kroenke, 1972; Shipboard Scientific Parties
of Legs 89 and 90, 1986; Kroenke, Berger, Janecek, et al., 1991). In
the Nauru Basin, immediately east of the plateau, sediment is <300
m thick. The slope of the flank of the plateau averages about 0.19°.
Seismic refraction studies show the crust of the plateau to be approximately 40,000 m thick (Kroenke, 1972; Furumoto et al., 1976;
Carlson et al., 1980). Crustal seismic velocities are on the order of
those of oceanic crust, but the thicknesses of crustal layers on the
Ontong Java Plateau are five times normal (Furumoto et al., 1976;
Hussong et al., 1979; Carlson et al., 1980).
Several aspects of the Ontong Java Plateau's history make the
plateau ideally suited for detailed paleoceanographic studies. First
and foremost is the combination of its bathymetry and geographic
location. The plateau, straddling the equator for much of its history,
is located in a region characterized by relatively high production of
biogenic sediment. The juxtaposition of the top of the plateau, at about
2000 m, and the flank of the plateau, down to 4500 m, requires that
pelagic sediment in both locations will have been produced in similar
surface-water conditions. Based on studies of benthic foraminifers,
the plateau seems to have remained at a relatively constant water
depth, at least since the Tertiary (Resig et al., 1976; Shipboard
Scientific Party, 1975). Most of the plateau has, therefore, remained
above the carbonate compensation depth (CCD) for at least the last
30 m.y. and has accumulated a thick pile of biogenous carbonate
sediment that has not been subjected to pervasive dissolution. Sediments along a depth transect on the flank of the plateau have been
exposed to increasingly deeper and more carbonate-corrosive waters.
The depth range of the flank (2000-4500 m) is precisely that through
which changes in dissolution gradients are most pronounced (Berger
and Johnson, 1976; Berger and Mayer, 1978). The combination and
constancy of bathymetry and geography of the Ontong Java Plateau,
34
therefore, eliminate many of the variables normally associated with
pelagic sedimentation (i.e., productivity and latitudinal gradients, and
tectonic bathymetric changes) and create a nearly ideal natural laboratory for evaluating the vertical distribution of a range of oceanic
parameters, both spatially and temporally.
Background
The main factors controlling pelagic carbonate sediment deposition
and accumulation are (1) the productivity of foraminifers and coccoliths in the overlying surface water; (2) the dissolution of carbonate
material as it falls through the water and lays on the seafloor; (3) the
dilution of carbonate particles by noncarbonates, such as terrigenous
particles, pelagic clay, volcanic detritus, or silica; and (4) the winnowing, scouring, and eroding of material on the bottom (Hamilton et al.,
1982), as well as other sedimentation processes (Berger and Johnson,
1976; Berger et al., 1977). Mayer et al. (1985,1986) have shown that
the major sedimentary response to paleoceanographic events in the
central equatorial Pacific is fluctuations in the amount of calcium
carbonate dissolution. They have shown that there are basin-wide,
seismically expressed sedimentary responses for paleoceanographic
events related to increased dissolution. Although discrete periods of
increased carbonate dissolution create events that are extremely useful
from a seismic or stratigraphic perspective, complete removal or severe
compression of the section that results from this dissolution makes
detailed evaluation of such paleoceanographic indicators as isotopic
signals, faunal changes, and chemical tracers virtually impossible at
these critical times. The Ontong Java Plateau provides a setting in
which Neogene and Quaternary sediment at the top of the plateau, in
2000 m of water, has undergone little carbonate dissolution, and thus
the section is relatively expanded. The flank of the plateau extends
through the present and past lysoclines and CCDs (van Andel et al.,
1975; Berger etal., 1977).
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
A seismic reflection profile, transecting the flank of the plateau,
can potentially yield a continuous record, in both spatial and temporal
senses, of the sedimentary column. Tied to drill sites to provide
valuable stratigraphic data, seismic reflection profiling becomes a
powerful tool for interpretation and extrapolation of the sedimentary
history of this region.
METHODS
Seismic Reflection
Three thousand kilometers of digital, single-channel seismic reflection data have been acquired from the study area on the Ontong
Java Plateau during Roundabout Leg 11 of the Thomas Washington
in December 1988 (Fig. 1) (Mayer etal., 1991). The seismic data were
collected using a 1.3 L (80 in.3), 136 atm (2000 psi) SSI water-gun
source. The water-gun source has a much broader bandwidth than
conventional air guns and, because it does not expel air, does not
produce bubble reverberations. A Teledyne streamer with 48 acceleration-canceling hydrophones in a linear array was used to receive
the acoustic signals. Depth was monitored with a calibrated depth
sensor at the head of the leader. Towing depth was held around 5 m
for both the gun and the streamer. Data were digitally recorded at
sampling intervals of 1 ms on nine-track magnetic tape with amplitude
preservation for later processing and display.
Single-channel seismic reflection data were also collected from
the JOIDES Resolution during Leg 130 (Fig. 1). The source system
consisted of two synchronized 1.3 L (80 in.3), 136 atm (2000 psi) SSI
water guns. The receiver consisted of a 100-m long Teledyne Model
178 streamer, towed at a depth of about 10 m. These data were also
logged at 1-ms sampling rates on magnetic tape (Kroenke, Berger,
Janecek, et al., 1991).
The Thomas Washington seismic reflection data consist of two
lines that run up the flank of the Ontong Java Plateau, tying into Deep
Sea Drilling Project (DSDP) Sites 289 and 586, plus six areas on the
flank that were surveyed in some detail as proposed (and later drilled)
ODP sites (Fig. 1). These data were analyzed at the University of
Texas Institute for Geophysics with SIOSEIS, a seismic processing
software package, and GEOQUEST, a seismic interpretation software
package. The interpretation software assists development of seismic
stratigraphy by facilitating reflection horizon tracing and providing
stratigraphic control at line cross-over points. The value of the system
has been the output of digitized, interpreted stratigraphic data, as
every trace is flagged with the depth-to-reflector horizon chosen in
the interpretation. The seismic stratigraphy data set, therefore, is
composed of thousands of data points, permitting accurate representation of the stratigraphy and statistically accurate averaging for
further calculations. It is these data that form the basis of the seismic
stratigraphy for this paper.
Physical Properties
Five sites on the northeastern margin of the Ontong Java Plateau
were drilled during Leg 130 (Sites 803-807). These sites form a depth
transect down the flank of the plateau (Fig. 1). In addition to continuous core recovery, downhole logging was conducted at all sites except
Site 804. Data from two sites on the top of the plateau from DSDP
drilling (Sites 289 and 586, 80 m apart) on Legs 30 and 89, respectively, are also available (Shipboard Scientific Party, 1975; Shipboard
Scientific Parties of Legs 89 and 90, 1986).
Physical properties at each of the sites were measured in situ with
downhole logging tools, and in the lab on discrete samples taken at
frequent intervals within the cores after splitting. Downhole logging
data pertinent to this study include compressional sonic velocity (both
near and far fields), and lithodensity (gamma-ray flux). Discrete
samples were taken from the cores every 0.75 to 1.50 m for saturated
bulk density, porosity, grain density, and dry-bulk density measurements (see Kroenke, Berger, Janecek, et al. [1991] for details of the
methods used to obtain these measurements). Longitudinal and transverse compressional sound velocities were measured directly within
the unlithified sediment of the cores using the digital sound velocimeter (DSV) (Mayer et al., 1987; Courtney and Mayer, in press) or, on
lithified sediment, with samples taken from the core, using the Hamilton frame (Boyce, 1973a, 1973b). Measurements were made every
0.75 to 1.50 m in the cores.
OBSERVATIONS
Seismic Reflection Data
The top of the Ontong Java Plateau is characterized by a thick
sediment column, on the order of 1 s (two-way traveltime) from
seafloor to acoustic basement, with numerous, closely spaced, parallel, continuous reflectors (Fig. 2, in back pocket). Many individual
reflectors can be correlated throughout the seismic profiles on top of
the plateau. Some reflectors are slightly higher in amplitude than
surrounding reflectors, though amplitudes may decrease laterally. The
distinction between high- and low-amplitude reflections become
more apparent toward the edge of the plateau. The seafloor topography is flat to slightly undulating, reflecting basement topography.
Acoustic basement is marked by several high-amplitude, discontinuous reflections underlying the sediment column.
The slope break between the plateau crest and its flank occurs at
a water depth of 2250 m, but no distinction in the echo character of
reflectors in the sediment column between the slope and the plateau
were observed until 2800 m. Between this water depth and about
3375 m, numerous vertical displacements and reflector disruptions
were observed, resulting in an incoherent seismic facies (Fig. 2).
These features are reflected in a rougher seafloor topography than on
top of the plateau. The sediment column thins by about 30% through
this depth range.
Reflectors become coherent again at about 3375 m water depth.
The echo-character of the sediment section changes significantly as
compared with the upper slope and plateau top (Fig. 2). Fewer
reflection horizons in the sediment column and changes in dip direction between reflections, resulting in thinning and pinchout, were
observed. Vertical displacements of reflections interpreted as faults
are common. Recent and buried channels, slump scars and deposits,
and fault displacements impart a rough seafloor topography.
On the lowermost portion of the slope, in water depths from 3800
to 4500 m, acoustic basement is extremely irregular with a horstand-graben-like topography (Fig. 2). The nature of this basement
morphology is thought to be fault controlled, possibly related to
differential subsidence between the plateau and adjacent normal
oceanic crust (Hagen et al., this volume). For conciseness, we will
refer to these small, local basins as grabens and to the high areas
between as horsts. Sediment reflectors are upturned and terminated
at the sides of the horsts, and the sediment section thins markedly
over them, making correlation of individual reflectors difficult.
Distinct seismic facies are recognizable within the sediment column
of the grabens.
Seismic reflection profiles in the Nauru Basin adjacent to the
Ontong Java Plateau show a sediment section on the order of 0.2 to
0.5 s thick overlying the reverberant layer forming acoustic basement
(Fig. 2). The horst-and-graben topography is not present in this
region. Sediment reflections pinchout and outcrop away from the
slope of the Ontong Java Plateau, making correlation of reflectors
over any lateral extent difficult.
Seismic Correlation
Eight reflecting horizons, including the seafloor and basement,
have been correlated throughout the available seismic reflection data,
including correlation to each ODP and DSDP drill site. Figures 3A
and 3B are two seismic stratigraphy interpretation profiles of the flank
of the plateau (see Fig. 1 for the locations of these profiles). Figure 3A
35
D.C. MOSHER ET AL.
A
West
East
50
100
150
200
250
300
350
400
450
500
Distance (km - along track)
B
Northeast
Southwest
2 2500
DSDP Sites 289/586
ODP Site 806
Seafloor
£
>. 3000 3500 "
o
— 4000 "
4500 "
5000-
H.5500
i
Φ
Q
-50
50
•
•
'
•
i
•
•
100
•
'
r
150
•
•
'
•
i
'
•
^
200
250
300
350
Distance (km - along track from DSDP Site 586)
Figure 3. A. InteΦreted seismic stratigraphy of the seismic section shown in Figure 2, extending from the top of the Ontong Java Plateau to its base in a west-east
direction (see Fig. 1). B. Interpreted seismic stratigraphy extending part way down the flank of the plateau in a southwest-northeast direction (see Fig. 1). Rl to
R6 represent the sediment reflection horizons correlated throughout the seismic data set. Vertical lines represent faults inteΦreted from the seismic data. Note the
slope break occurs in about 3000 ms (2250 m) water depth, but significant sediment column thinning begins at about 3700 ms (2800 m) water depth. These plots
were generated with digital data, output from GEOQUEST (a seismic inteΦretative software package). Each horizon represents about 2000 data points (every
fifth trace of the seismic data).
is the interpretation of the seismic profile shown in Figure 2 (in back
pocket). Bathymetry of the plateau flank dips and then rises again
before plunging into the Nauru Basin. A sediment thickness plot
(Fig. 4) shows that the sediment column thins most significantly
over this rise and then continues to thin downslope. At the base of
the flank, sediment thicknesses vary significantly because of infilling in the grabens. Figures 5 to 10 show the seismic section crossing
and seismic stratigraphy at each of the drill sites. Figure 11 is a
sediment thickness vs. water depth plot, summarizing the seismic stratigraphic data. From this figure, one can see that the sediment
thins with increasing water depths. It shows that the entire sediment
section has thinned 65% through these water-depth ranges (i.e., it is
35% as thick at the base of the plateau as at the top), and that thinning
is most significant below 2900 m water depth. It is difficult to tell,
36
however, from this figure where in the sediment column this thinning
is occurring. Depths must first be converted to meters from traveltimes for calculations of the amount of thinning to become significant. This conversion can be accomplished through modeling the
seismic response of the sediment column, which requires sediment
physical properties information.
Synthetic Seismograms
Physical Properties
The two sediment physical properties that this paper is concerned
with are compressional wave velocity and sediment bulk density. The
product of these two properties is acoustic impedance; changes in
impedance reflect sound, creating seismic reflections. Velocity and
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
I. . . . I. . . . I. . . . i. . . . i
3000-
2000
-1800
-1600
-1400 |
Seafloor bathymetry
5000-
eooo-
-1200 |
-1000 α
"800
|
"600
1
Sediment thickness
400
200
7000
0
158
159 160 161 162 163 164 165
Longitude (east)
Figure 4. A profile of seafloor bathymetry (left-hand vertical axis) and sediment
thickness (right-hand vertical axis), on the flank of the plateau, generated from
the digital seismic stratigraphy data. Data are plotted with longitude as the
.t-axis, ignoring latitude; thus, true distance along track is not apparent. The
data, perpendicular to the contours of the slope, are close enough to parallel a
latitude line that a 15%-weighted, smoothed curve through the data shows the
profile, and the sediment thickness of the plateau, well. Note that the most
extreme sediment thinning occurs between 2900 and 3225 m water depth.
density data were collected by discrete measurements made in the
laboratory from recovered cores and through downhole logging with
sonic velocity and gamma-ray density tools. Logging data are assumed to represent in-situ conditions. Laboratory data must be corrected to in-situ conditions of pressure and temperature.
Density Corrections
Removal of overburden pressure by the extraction of sediment
core from in situ to the laboratory causes an increase in porosity in
unlithified sediments (rebound porosity). A bulk-density measurement made on core material, therefore, must be corrected for porosity
rebound. Porosity rebound can be simulated in a laboratory with a
standard civil engineering test known as a consolidation test (Hamilton, 1971b; Das, 1983; Marsters and Manghnani, this volume). Marsters and Manghnani (this volume) conducted 19 of these tests on
samples collected during Leg 130 and formulated the porosity rebound relationship for Ontong Java Plateau sediments as follows:
% rebound = 4.67 × 10"3D - 1.97 ×
(1)
where D is depth in meters below seafloor.
The saturated bulk density of a sample is governed by the following relationship:
where <j) = porosity, p^ = fluid density (seawater), and pg = grain
density (carbonate).
Correction of porosity to in-situ conditions with the porosity
rebound equation then allows one to correct the bulk density to in-situ
conditions. Corrected values of bulk density compare favorably to
gamma-ray density values measured downhole (Fig. 12). The correc-
4.0
-
Figure 5. Seismic reflection profile on the Ontong Java Plateau, crossing over
DSDP Sites 289 and 586. The data comprising this profile were collected on
the Thomas Washington during Roundabout Leg 11. Note the lateral continuity,
similar amplitude, parallel, gently undulating nature of the reflections, typical
of the top of the plateau. The ship track turns over the drill site; hence, south
is in the middle of the section (see Fig. 1, DSDP Site 289/586 crossing). Rl to
R6 are the reflection horizons correlated throughout the seismic data set (see
Fig. 3B). The synthetic seismogram generated from the impedance profile is
shown in the middle of the section at the drill site.
tion, in fact, proves to be quite small (2.7% decrease in porosity at
1000 m below seafloor).
Velocity Corrections
Compressional sound velocity in an isotropic material is a function
of elastic properties and saturated bulk density of the rock/sediment,
according to the following relationship (Wood, 1941; Hamilton, 1970):
K + 4/3µ
V2 =
Psat
(3)
where V= sonic velocity, K = bulk modulus, µ = shear modulus, and
psat = saturated bulk density.
Mayer et al. (1985) have shown that correction of velocity to
in-situ conditions in the central equatorial Pacific is possible by the
relationship between laboratory velocity and laboratory porosity.
Correction for porosity rebound merely shifts the curve, and corrected
velocity values can be predicted from the adjusted curve. For the
sediments of Leg 130, however, a good relationship between porosity
and compressional velocity does not exist, implying that compressional velocity is not solely dependent upon porosity (Fig. 13) (Urmos
et al., this volume). Porosity, therefore, could not be used to predict
velocity. Hamilton et al. (1982) similarly found a poor relationship
between compressional velocity and porosity and density for box core
sediments on the Ontong Java Plateau. Other methods are required,
therefore, to correct velocity to in-situ conditions.
37
B.C. MOSHER ET AL.
3.5
3.7
"― 3.9
CD
E
"03
2 4.1
+-<
>,
CO
4.3
4.5
4.7
Figure 6. Seismic reflection profile on the Ontong Java Plateau crossing over
ODP Site 806. These data were collected on the Thomas Washington during
Roundabout Leg 11. This site is located on the uppermost part of the plateau
flank in 2521 m of water (see Figs. 1 and 3A-3B). Note the similarity of the
section with that of Sites 289/586 (Fig. 5). Rl to R6 are the reflection horizons
correlated throughout the seismic data set. The synthetic seismogram is posted
in the middle of the figure at the drill-site position.
The first correction to compressional velocity is for change in density,
as shown above, as density appears in the denominator of Equation 3.
Velocity must also be corrected to in-situ conditions of pressure and
temperature. As shown by Wyllie et al. (1956) for rocks with rigidity,
velocity is a function of pore fluid velocity and grain velocity:
l J
V
(1-Φ)
vo
(4)
where Vb = bulk velocity, Vw = velocity of pore fluid (seawater), Vg =
velocity of mineral grain, and Φ = porosity.
Boyce (1976) developed the following correction for velocity to
in-situ conditions in carbonate sediment of the Pacific Ocean, by
assuming that the ratios of respective velocities were constant in the
lab and in situ.
Vm =
(5)
where V{ = velocity of seawater at lab conditions, V2 = velocity of
mineral grain (calcite) at lab conditions, V3 = velocity of sample at
laboratory conditions, V7 = velocity of seawater at in-situ temperature
and hydrostatic pressure, Vπ = velocity of calcite at in-situ temperature and hydrostatic pressure, and Vm = velocity of sample at in-situ
temperature and hydrostatic pressure.
The velocity of seawater can be calculated based on temperature,
salinity and hydrostatic pressure using equations developed by Wilson (1960) derived from measured sound speed data. The temperature
downhole is calculated using a geothermal gradient of 0.044°C/m.
This value was derived from heat flow measurements on the Ontong
Java Plateau reported in Langseth et al. (1971). The velocity of calcite
38
-
Figure 7. Seismic reflection profile on the Ontong Java Plateau crossing over
Site 807, in 2804 m water depth (see Fig. 1). These data were collected on the
JOIDES Resolution during Leg 130. The data are of lower frequency content
than the Thomas Washington data, making it difficult to correlate with the
Thomas Washington seismic stratigraphy. To the south and north of Site 807,
the sediment thickness is the same as on top of the plateau and reflections are
roughly similar in character to those of the top of the plateau. Site 807, however,
is located in a depression in acoustic basement, corresponding to a thicker
sediment section. Some disruptions in reflections are associated with this
basement low. Rl to R6 are reflection horizons correlated with the traced
reflection horizons from the Thomas Washington seismic data. The synthetic
seismogram is posted in the middle of the section at the position of the drill
site. The seafloor return from the field seismic data was used as the source
wavelet in the convolution to generate this seismogram, because the seismic
source wavelet was obviously significantly different from the source used
during the Thomas Washington survey.
is assumed to remain constant with depth and temperature at 6450 m/s
(Boyce, 1976).
Application of these corrections to laboratory-measured velocities
produces a downhole curve of velocity vs. depth that shows significantly lower velocities than the logging velocity curve (Fig. 14). In
other words, the laboratory-measured velocity, corrected by the methods suggested by Mayer et al. (1985) and Boyce (1976), do not
compensate for in-situ conditions, in this depositional environment,
when compared with logging velocity data. We think that this disparity is likely a function of elastic moduli that appear in velocity
Equation 3. Bulk moduli (K) and shear moduli (µ) are not accounted
for in the above corrections. That is to say, the ratios of velocities are
not constant as Boyce (1976) assumed in his corrections. One would
expect that increasing overburden pressure would tend to increase
values of K and µ, as it would tend to enforce sediment grain-to-grain
contacts. Urmos et al. (this volume), in attempting to correct velocity
data to in-situ values, have reached many of the same conclusions.
Rigidity effects on velocity are especially important in the
foraminifer-rich sediment considered in this study, as Fulthorpe et al.
(1989) also noted in their seismic studies of the Ontong Java Plateau.
Foraminifer tests are hollow spheres with intraparticle porosity, so the
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
4.0
4.5
4.2
4.4
i-
4.7
CD
j | 4.9
CD
4.6
>,
£0 5.1
ó
I 4.8
5.0
5.2
Figure 8. Seismic reflection profile on the Ontong Java Plateau crossing over
Site 805. These data were collected on the Thomas Washington during Roundabout Leg 11. The site is located in 3188 m water depth (see Figs. 1 and 3B).
The sediment column has thinned 25% compared with Site 806. The reflection
profile is similar in character to sections higher on the plateau, although some
reflections are of slightly higher amplitude. R2 to R6 are the correlated
reflection horizons, traced throughout the seismic data set. Some evidence of
faulting is present, noted by a displacement of reflectors within the sediment
column. A synthetic seismogram is posted in the middle of the figure at the
drill site position.
nature of the grain-to-grain contact can be changed without significantly decreasing total porosity. For example, experiments by
Laughton (1957), in which he simultaneously measured consolidation
and velocity on foraminifer ooze, found that a 5% change in porosity
resulted in a 64% change in velocity (3.05 to 1.96 km/s), which he
mainly attributed to a change in grain-to-grain rigidity (shear modulus). Hamilton et al. (1982) also suggested that, for the Ontong Java
Plateau, shear rigidity (µ) was the dominant cause of velocity variations because the density (p) and bulk modulus (K) terms of Equation
3 tend to cancel. Corrections to elastic moduli, however, cannot be
readily applied because they vary significantly between sediment
types and geologic setting and because little information is available
on elastic and viscoelastic properties of water-saturated, uncemented
marine sediments at low pressures (see Hamilton, 1971a, 1971b,
1972, 1974, 1980).
To adjust for differences observed in lab and log velocity data,
laboratory measurements within the ooze interval were further corrected to logging velocity curves in the ooze interval. This correction
is accomplished by determining a best-fit linear regression curve
through logging velocity and lab velocity data (Fig. 15). The regression through the log data is forced through the same point of origin
as the lab data regression because at the seafloor no rigidity effects
exists (and the lab data is already corrected for temperature). Lab
velocity data is then adjusted so the trend of lab values is equivalent
to the trend of log velocity values. A different curve was calculated at
each drill site to account for differing sediment characteristics at each
site. The final corrected laboratory compressional wave velocity
measurements are assumed to represent in-situ conditions (Fig. 16).
One should note that logging compressional velocity values are
not considered to be correct for the ooze intervals, as the coring
process tended to enlarge the diameter of the drill hole beyond the
5.3
5.5
Figure 9. Seismic reflection profile on the Ontong Java Plateau crossing over
Site 803. These data were collected on the Thomas Washington during
Roundabout Leg 11. The site is located in 3410 m water depth (see Figs. 1
and 3A). The sediment column has thinned by about 40% compared with the
top of the plateau. The profile shows parallel reflectors at the site. Higher
amplitude reflections can be seen in the middle of the section and a mid-section reverberation (MSR) layer is observed next to the site. These features
are common within the sediment section at about this water depth on the flank
of the plateau and have been observed elsewhere throughout the Pacific
(Houtz and Ludwig, 1979; Shipley et al., 1985). High amplitudes of reflection
wavelets comprising these features suggest high impedance contrasts with
surrounding sediment. They have been interpreted as volcanic sills and as
diagenetic horizons (Kroenke, 1972; Houtz and Ludwig, 1979; Shipley etal.,
1985). In this case, the MSR occurs at the exact horizon of a Miocene hiatus
(between the R3 and R4 reflection horizons). An ash layer was also observed
at about the same depth as the MSR, though the relationship between the two
is uncertain. R3 to R6 are the reflection horizons correlated throughout the
seismic data set. A synthetic seismogram is posted in the middle of the section
at the drill site position.
accurate range of the transducers mounted on the logging string. In
other words, the velocities measured with the logging tool integrate
velocities of the water surrounding the tool as well as the sediment.
Through much of the ooze interval, the diameter of the holes is not
known because the holes are wider than the maximum extension of
the caliper tool. A power spectral density plot of logging />-wave
velocity, and gamma-ray density, through the ooze interval at Site 806
shows extreme dominant power at a wavelength of 9.5 m and two of
its harmonics (4.75 and 3.17 m) (Figs. 17A-17B). This periodicity is
likely a coring artifact, as it is precisely the length of the advanced
hydraulic piston core (APC). The power spectrum suggests that this
frequency component would completely mask any real components
in any mathematical operation that involved this data.
This coring disturbance is most pronounced at Site 806, but it is
also recognized at each Leg 130 site. In spite of these inaccuracies
in the high-frequency component of the data, we think that the data
trends are accurate and represent in-situ conditions, and it is these
trends to which the laboratory-measured velocity data are being
corrected. We also think that the corrected laboratory data of both
velocity and density are more accurate than the logging data over
the ooze interval. These data, therefore, are used for ooze intervals
and are merged with logging data at the ooze/chalk transitions
(between 240 and 350 mbsf). Downhole velocity and density profiles are, therefore, sampled with discrete measurements with a
maximum sample spacing of about 0.75 m for the laboratorymeasurement portion, and 0.15 m for the log-measurement portion
39
D.C.MOSHERETAL.
Figure 10. Seismic reflection profile on the Ontong Java Plateau crossing over
Site 804 collected on the Thomas Washington during Roundabout Leg 11. This
site is located in 3862 m water depth, within a basement depression, thought
to be a graben structure (Hagen et al., this volume). The sediment section is
about half as thick as on top of the plateau. Reflections vary greatly in amplitude
with depth, and correlation outside of the depression is difficult. Evidence of
active sedimentation exists in this isolated basin, with onlap and pinchout of
the reflections against its walls. Rl to R6 represent the correlated reflector
horizons. No downhole logging was conducted at this site, so no data exist for
generation of a synthetic seismogram.
of the curves. The product of velocity and density is impedance; thus,
these profiles provide the basis of the reflection coefficient series for
seismogram modeling.
Correlation to Seismic Data
It is necessary to correlate Leg 130 drill sites to the acoustic
stratigraphy to examine the response of the seismic signal to changes
in physical properties within the sediment column. The most critical
and difficult step in applying strati graphic significance to seismic data
is the transformation of sub-bottom depth information to traveltime
information, which requires an exact knowledge of the in-situ velocity. This domain transformation can be accomplished through the
generation of synthetic seismograms.
Synthetic seismograms model the interaction between a seismic
wavelet and a given well-constrained geologic model. With an accurate representation of the seismic source wavelet and an accurate and
detailed picture of subsurface geologic variations, the synthetic seismogram should provide a record of their interactions. The geologic
model is the EartiYs reflection series, which is essentially the change
in acoustic impedance with depth, impedance being the product of
compressional wave velocity and bulk density. The synthetic seismogram is produced by mathematically convolving the source function
(in this case the signal produced with a 1.31 L water gun) with the
reflection coefficient curve (change in impedance) at the site.
At each drill site on the Ontong Java Plateau, a synthetic seismogram was generated from merged log and lab velocity and density
data, except for Site 804, for which no logging data exists. Analyses
of the field data showed that the source function was of lower
frequency content (0-100 Hz) than the calibrated hydrophone test
performed during the same cruise that produced the source function
used in the convolution (0-300 Hz). As a consequence, the synthetic
seismograms were bandpass-filtered (0-100 Hz) to approximate the
frequency content of the seismic data. Figures 5-10 show the results
of synthetic models with correlation to the seismic reflection data over
each site. We also found that a static shift in velocity values of -30
m/s was required to accomplish the appropriate traveltime-to-depth
40
conversion for horizontal correlation of the synthetic seismogram to
field seismic data. This correction implies that the logging velocity
tool overestimates true in-situ velocity by 30 m/s.
Figure 18 is a cross-plot of the synthetic-to-seismic correlation
results shown in Table 1. This figure shows that correlations at each
site all fall on a single curve, implying that a single velocity-depth
function applies to all the sediments of the Ontong Java Plateau,
regardless of the water depth in which they were deposited, or their
age, or state of induration. The equation of a second-order polynomial
curve fit through this data allows the conversion of traveltime on the
seismic section to depth in meters below seafloor, or vice versa.
Figures 19Aand 19B show the results of site-to-site correlations based
on the seismic stratigraphy and synthetic seismogram depth conversions, as compared with results of site-to-site correlation based on an
inverse-signal correlation technique of logging density and resistivity
curves (Lyle et al., this volume). These data of Lyle et al. (this volume)
provide a means of checking seismic stratigraphic correlations and
the conversion of traveltime to depth. The results of these comparison
plots show reasonable agreement between the correlations of Lyle et
al. (this volume) and those based on seismic stratigraphy.
DISCUSSION
Three striking changes occur in the character of seismic profiles
from the top of the Ontong Java Plateau (2000 m water depth) to the
Nauru Basin (4500 m water depth): (1) a drastic decrease in sediment
thickness (about 75%); (2) the loss of continuous, parallel reflectors;
and (3) development of distinct seismic facies (i.e., intervals in
profile of high-amplitude, continuous reflections, overlying or underlying intervals of lower amplitude or incoherent reflections) with
increasing water depths, as opposed to the top of the plateau, which
demonstrates one sequence of similar echo-character throughout the
sediment column.
The seismic character of low-amplitude, laterally continuous, and
largely parallel reflectors across the top of the Ontong Java Plateau
and down to a water depth of about 2800 m water depth are consistent
with the uninterrupted blanket of post-upper Oligocene pelagic sediments described for Sites 289, 806, and 807 (Shipboard Scientific
Party, 1971; Kroenke, Berger, Janecek, et al., 1991). The cause (or
causes) of this echo-character is not entirely clear. Mayer et al. (this
volume) interpret grain-size fluctuations as the dominant factor in
determining changes in bulk density as observed in gamma-ray
attenuation and porosity evaluator (GRAPE) data. The grain-size
distribution naturally affects the nature of grain-to-grain contacts. We
expect, therefore, that sediment grain size also greatly influences
compressional velocity. Because grain size can affect density and
velocity, then grain-size fluctuations may account, at least in part, for
seismic reflections.
In the pelagic environment of the Ontong Java Plateau, sediment
grain-size distribution can be affected in two ways: (1) productivity/preservation conditions, in which the size distribution is determined by the sizes of tests produced and the amount of selective
dissolution that has affected these particles; and (2) resedimentation,
in which pelagic sediment is syndepositionally or postdepositionally
affected by sedimentation processes, namely, winnowing on the plateau top. Wu and Berger (1991) and Mayer et al. (this volume)
consider winnowing to be an important process on top of the plateau,
related to paleoceanographic events. Aside from a slight thickening
of seismic units toward the edges of the plateau, which will be
discussed below, there is little evidence in the seismic reflection data
for large-scale winnowing. Reflectors on top of the plateau are parallel
and continuous over long distances (hundreds of kilometers). If
winnowing has been an important process, it has not disrupted the
sediment column, produced any visible bedforms, or significantly
eroded sediment on top of the plateau. The seafloor topography
largely reflects basement morphology, in spite of the fact that the
sediment column is over 1000 m thick. Seismic evidence, therefore,
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
0 Φ
1
200 -
400 H
CO
ε. eoo 800 -
1000 -
1200
2000
2500
3000
Water depth (m)
3500
4000
4500
Figure 11. Sediment thickness vs. water depth. This plot shows the thickness from the seafloor to each of the reflection horizons at given water depths.
The entire digital seismic stratigraphy data set was used to produce these plots, not just a single line; hence, the scatter of points. A 15%-weighted,
smoothed curve was used to fit the data for each horizon.
would support the theory that productivity/preservation changes are
the mechanisms responsible for grain-size fluctuations on top of the
plateau rather than sedimentation processes.
Evidence exists, however, within the seismic reflection and core
data that resedimentation processes are active on the flank of the
Ontong Java Plateau. Evidence of faulting and mass wasting that has
been interpreted from the recovered sediment (Kroenke, Berger, Janecek, et al., 1991) can account for loss in reflector continuity, and the
development of seismic facies, on the flank of the plateau. This
evidence is especially apparent in the incoherent section between 2900
and 3300 m water depth (Fig. 2), which coincides with the steepest
slope and the greatest amount of sediment thinning on the flank.
Site 807 is located in 2800 m water depths within a slight basement
depression. Seismic reflection profiles at the site show reflector
displacements and incoherent reflections within the sediment associated with this downdrop in basement. Sediment cores demonstrate a
number of intervals within which microfaults are common, from as
shallow as 49 mbsf to the base of the sediment section (Kroenke,
Berger, Janecek, et al., 1991). These features are likely associated with
reflector discontinuities in the seismic profile. Below about 1116 mbsf
(lower Eocene), numerous intervals bearing suspended clasts up to
0.05 m in diameter were observed. These clasts typically contain older
nannofossil assemblages than the matrix (Kroenke, Berger, Janecek,
et al., 1991). These zones are thought to represent erosion of outcropping material further upslope and transport into the basin by debrisflow processes. Overlying these clast-bearing sediments is a
35-m-thick limestone interval containing well-sorted foraminifer
sand layers interpreted as possible turbidites. In addition, most of the
claystones and radiolarian siltstones that overlie basement are interpreted as turbidites (Kroenke, Berger, Janecek, et al., 1991). Seismic
stratigraphy analyses show that the unit between the R6 horizon and
acoustic basement is greatly overthickened at this site relative to Site
586 (Fig. 20).
There is evidence also for active resedimentation processes
throughout much of the flank of the plateau. Thinning of the sediment
column with depth can be explained by increased carbonate dissolution with water depth. Drilling results have shown that the sediment
column at each of the sites down the flank of the plateau is composed
of 80%-90% carbonate material (Kroenke, Berger, Janecek, et al.,
1991). Experiments by Berger (1967, 1968, 1970), Wu and Berger
(1991), and Wu et al. (1991) have demonstrated that the present-day
lysocline and CCD of the Pacific Ocean and of the Ontong Java
Plateau in particular, lie at about 3500 and 4000^500 m water depths,
respectively. If present conditions have prevailed over the last 40 m.y.,
then the sediment thickness would remain constant with depth down
to about 3500 m. It would drop off suddenly at 3500 m and, by
definition of the CCD, should be nonexistent below 4500 m (i.e., no
carbonate preservation). The profiles of sediment thickness (Figs. 4
and 11), however, show that the sequence begins to thin significantly
at about 2900 m water depth (3900 ms). The sediment section continues to thin but is still about 25% as thick as the top of the plateau
by 4500 m water depth. It is still composed of about 70% carbonate
material at this depth (Berger et al., 1977). Differences in this profile
from the theoretical described above suggests that either the lysocline
and CCD were not in the same position in the past as they are at
present, and/or the profile has been significantly altered by masswasting events. Sediment failure would transport material from the
upper portion of the plateau flank to below the CCD. Rapid deposition
resulting from mass failures also would ensure isolation of sediment
from the water column to prevent significant sediment dissolution. In
this way, mass-wasting processes modify sediment distribution down
the flank.
Evidence exists in the seismic reflection profiles from most of
the plateau flank below a water depth of 2800 m for scarps, faults,
erosional channels, slump blocks and reflector pinchouts, and discontinuities indicative of mass-movement events. Though drilling
sites were chosen in an attempt to avoid such features, sediments
recovered at Sites 805 and 803 exhibit evidence of microfaulting
(Kroenke, Berger, Janecek, et al., 1991), which may be attributed to
larger scale faulting.
The fact that sediment failures are in the depth range between the
lysocline and the CCD suggests that dissolution may have played a
role in their genesis. It is possible that carbonate material deposited
on the slope below the lysocline is removed or weakened through time
41
D.C. MOSHER ET AL.
Density (g/cm3)
U
15
Uβ
1.7
100 -
* *
1500"
Linear fit
y-1930.7 -6.0193X
^-0.56
1480
72
70
63
J
64
62
Porosity (%)
60
53
56
Figure 13. Plot of laboratory measured compressional velocity (P-wave velocity) vs. porosity for Site 806 (0-350 mbsf). The poor correlation coefficients
(r 2 = 0.56 for a linear fit and 0.64 for a second-order polynomial fit) show that
velocity depends very little on porosity. 'The deeper sites show even poorer
correlation between these parameters. This plot is used as an example to show
that porosity cannot be used to "predict" velocity, to find a correction for
velocity to in-situ conditions for Ontong Java Plateau sediments, as Mayer et
al. (1985) did in the central equatorial Pacific.
£• 400 -
700 -
800
Figure 12. Comparison plot of uncorrected laboratory density data, corrected laboratory density data (using Marsters and Manghnani [this volume]
correction for porosity rebound), and logging gamma-ray density vs. water
depth. The lines are 5%-weighted, smoothed curves through the data. The
correction proves to be small but accurate, assuming the logging values
represent in-situ conditions.
42
by dissolution. This process removes support for sediment further
upslope, causing sediment failures. This dissolution and removal of
support is known as dissolution undercutting, as described by Berger
and Johnson (1976). Earthquake ground accelerations, which occur
frequently with the collision of Ontong Java Plateau and Solomon Arc
on the western side of the plateau (Kroenke, 1972; Kroenke et al.,
1986) further add to the frequency of sediment mass failures and may
be the triggering mechanism for sediment that has undergone dissolution undercutting.
Toward the base of the plateau flank (i.e., 3800^500 m water
depth) strong evidence for resedimentation processes is apparent on
the seismic reflection profile. The sediment column is complicated by
the rough acoustic basement morphology, comprised of a series of
small basins interpreted as grabens (Hagen et al., this volume) (Fig. 2).
Reflectors truncate and upturn at the sides of these basins, probably
resulting from either differential compaction or sediment offlap or
onlap during deposition. The truncation of reflectors at the surface
above basement highs suggests that seafloor erosion is taking place
and the depocenters are these small grabens. Shipley et al. (1989) argue
that basin filling is an active sedimentation process in the deep sea of
the equatorial Pacific. Basin filling comprises particle-by-particle
downslope movement of sediment occurring syndepositionally, as well
as postdepositional erosion, downslope creep, slumping, and local
turbidity current processes that move sediments from the slopes and
highs into local lows (Johnson and Johnson, 1970; Moore and Heath,
1967; Mudie et al., 1972). Basin filling produces deposits characterized
by onlapping onto surrounding highs (Shipley et al., 1989).
Site 804 was located within one of these basement grabens
(Fig. 10). Kroenke, Berger, Janecek, et al. (1991) reported that the
lower Pliocene contained mixed assemblages of older fauna, indicating sediment reworking; numerous dipping color bands with associated microfaults were observed; sorted and graded beds interpreted
to be turbidites were apparent in the late and middle Miocene; and
between 187 and 207 mbsf, an interval of highly disturbed sediment,
interpreted as a slump deposit, was noted. This slump interval corresponds with a hiatus between 197.1 and 206.8 mbsf in the early to
middle Miocene (14.7-20.8 Ma).
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
Velocity (m/s)
1400
1600
i
1800
i
2000
i
2200
i
2400
i
2600
2800
i
100-
200 "
300
<Λ
S3
ε400
f
3
500
600 "
700 -
800
Figure 14. Plot of unconnected and corrected laboratory F-wave velocity
(corrected for density changes resulting from porosity rebound and in-situ
conditions of pressure and temperature) and logging velocity vs. depth below
seafloor. This plot is used as an example to show that even with corrections,
laboratory velocity do not accurately represent assumed in-situ conditions.
On the extreme lower end of the flank of the Ontong Java Plateau
and into the Nauru Basin, seismic reflection data show distinct
differences in reflection patterns (Fig. 2). Reflectors tend to pinch out
and truncate against the seafloor. We interpret that the appearance of
the reflection profile is a result of active sedimentation, particularly
turbidity current activity, bringing material from the plateau into the
basin. The fact that carbonate material exists here at all, below the
CCD, would support this interpretation.
Although faulting, mass wasting, and other resedimentation
processes can account for reflector discontinuities and the development of seismic facies, and they can account for the redistribution of
sediment, they cannot account for the degree of sediment thinning
with increasing water depth down the flank of the plateau, which is
the most striking feature of the seismic reflection data. Thinning is
caused by the dissolution of carbonate material below the lysocline.
It is, therefore, dependent upon the position of the lysocline (water
depth) and on the dissolution gradient between the lysocline and the
CCD. Because sediments of the Ontong Java Plateau are composed
almost entirely of carbonate material, their distribution is sensitive to
the lysocline position, the dissolution gradient, and changes through
time of these parameters.
Figures 3A-3B, 4, 11, and 20 and Tables 1 and 2 show where and
when thinning is taking place in the sediment column. In Figure 20,
the greatest degree of thinning, observed at Sites 803 and 805, appears
to have occurred between the R3 and R4 horizons (middle Miocene),
and between the R5 and R6 horizons (Oligocene-early Miocene). The
R3 to R4 thinning corresponds to a Miocene hiatus (15.4-21.8 Ma)
at Site 803, and the R5 to R6 thinning correlates with a condensed
section in the late Oligocene (23.7-28.2) at Site 803 (Kroenke, Berger,
Janecek, et al., 1991). In this way, seismic reflection data are reiterating what is observed from the core evidence.
It is interesting to note that the sediment column between the Rl
and R5 reflectors (early Miocene to Pliocene) is thicker at Site 806
relative to Site 586, and is thicker between the R2 and R3 reflector
(late Miocene) at Site 805, relative to Site 586. Reflector correlations
to these sites have been accomplished with confidence and agree
reasonably well with the correlations of Lyle et al. (this volume).
Given that the correlations are correct, this apparent overthickening
can be explained by one of two mechanisms:
1. A slight latitudinal gradient is present between Site 586 at
0.29°S, Site 806 at 19'N, and Site 805 at 1°14'N. It is possible that
this latitudinal gradient corresponds to a latitudinal surface-water
productivity gradient. At present, the most productive water is
slightly offset from the equator at about 2°N (Berger et al., 1988;
Berger, 1989).
2. The second explanation for the overthickening involves winnowing processes that might have removed sediment from the shallowest parts of the plateau to be deposited at the top of the flank of
the plateau.
Regardless of the mechanism for the increased supply to the
deeper sites, the effects, we think, are dampened at the deeper sites
because of carbonate dissolution and the position of the lysocline
(Fig. 20). Lyle et al. (this volume) argue that other factors besides
carbonate dissolution are responsible for sediment accumulation
rates. Their results are much higher resolution than ours and are
complicated, therefore, by other sedimentation processes. From Figure 20, if we assume that sediment accumulation relative to the top
of the plateau is entirely a function of carbonate dissolution, we can
attempt to track the position of the lysocline through time. The trends
in our data at the crude resolution of the seismic stratigraphy are
consistent from site to site (Fig. 20). Thus, we think that a shallow
lysocline in the early Miocene caused sediment thinning, especially
evident at Site 803, but apparent even at water depths as shallow as
Site 806 (2500 m). A deepening of the lysocline occurred in the late
early Miocene, causing relative thickening at each site. Within the
43
D.C. MOSHER ET AL.
Velocity (m/s)
1400
1500
1600
1700
1800
1900
2000
i
i
i
i
i
i
If>
2100
Laboratory velocity
y = 1521.4+ 0.30x
Logging velocity
y = 1521.4 +1.28 x
so -
100-
150-
200-
middle Miocene, between the R3 and R4 horizons, a sharp rise
possibly took place in lysoclinal depth as well as a steepening of the
dissolution gradient, resulting in sediment thinning at all four sites,
though only slightly apparent at Site 806. The effects of this increased
dissolution can be seen in sediments of this age at Site 805, with a
decrease in abundance of whole foraminifer as compared with sites
on top of the plateau (Kroenke, Berger, Janecek, et al., 1991). Enhanced dissolution can reduce the relative importance of foraminifers
in sediments because foraminifer preservation is more sensitive than
calcareous nannofossil preservation to the position of the calcite
lysocline (Berger and Johnson, 1976; Berger et al., 1977; Berger and
Mayer, 1978). The thicker sections between the R2 and R3 horizons
(late Miocene) likely correspond to a deepening of the lysocline. A
subsequent rise in the lysocline again hinders accumulation of sediment in the very late Miocene and Pliocene. This rise results in some
thinning of the section between the seafloor and the R2 horizon at Site
805 relative to Site 586.
Figure 21 is an interpretation synthesis of the data presented in
Figure 20, estimating lysoclinal depths from the amount of seismicstratigraphic thinning that has occurred at each site, assuming again
that thinning is entirely related to carbonate dissolution. The trends
do not fit particularly well with the CCD curve presented for the
central equatorial Pacific (van Andel et al., 1975) (Fig. 21), although
Farrell and Prell (1989) show that changes in the depth of the
lysocline do not necessarily correlate with CCD changes. Van Andel
et al. (1975) show only slight changes in the CCD through the
Neogene, with a rise in the CCD in the early Miocene, and a slight
lowering again in the middle Miocene. We would predict a shallow
lysocline based on the degree of thinning in the middle Miocene.
Similarly, the deepening of the lysocline we would predict in the late
Miocene is not apparent on the CCD profiles (Fig. 21). These profiles
also show a static to decreasing trend in the depth of the CCD through
the Pliocene, whereas we would predict a shoaling of the lysocline.
These results imply that the lysocline fluctuations do not follow
CCD fluctuations, similar to conclusions of Farrell and Prell (1989),
or that sediment thinning is driven by mechanisms other than carbonate dissolution.
SUMMARY AND CONCLUSIONS
250-
300-
350
Figure 15. Plot of logging P-wave velocity and laboratory P-wave velocity
(corrected for in-situ temperature, pressure, and density) vs. depth for the ooze
interval at Site 806. The solid lines are the linear best-fit curves through the
data. The linear fit through the logging data is forced through the same
Y-intercept as the lab data, as velocity at the seafloor should be equivalent. The
lab-velocity data is corrected to the trend of the logging velocity data to
simulate assumed in-situ conditions.
44
The flank of the Ontong Java Plateau is an ideal setting to study
carbonate sedimentation through a depth transect of carbonate-neutral water to carbonate-corrosive water. A study of these sediments
with seismic reflection profiles, in collaboration with sediment
stratigraphic data acquired through drilling, provides a powerful way
to study the sedimentary history of the region in both spatial and
temporal terms. Over 3000 line-kilometers of single-channel seismic
reflection data and data from seven ODP drill sites on top of, and
along the flank of, the Ontong Java Plateau comprise the data base
for this study.
Results from ODP drilling include downhole compressional
velocity and sediment bulk-density measurements. Laboratory measurements of these parameters on cores recovered from drilling had to
be corrected to in-situ conditions. Density was corrected with a
porosity rebound curve generated by Marsters and Manghnani (this
volume); the correction is small (2.7% porosity at 1000 mbsf). Correction of velocity to in-situ conditions of pressure, temperature, and
density (porosity rebound) does not entirely compensate the values
to in-situ levels based on well-log results. We think thatrigidity,which
is largely dependent upon the depth of burial and grain-size distribution, is a significant component of the velocity formula in the case for
carbonate sediments of the Ontong Java Plateau. This elastic parameter is difficult to measure and account for in soft sediment, so the
laboratory-measured velocity has been further corrected to levels of
downhole logging sonic velocity values, which are assumed to represent in-situ conditions. The corrections to velocity are significant.
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
Velocity (m/s)
1400
1600 1800 2000
i
i
i
2200
2400
2600
i
i
i
2800
100 -
10
Wavelength (m)
200 -
Lab-log merge
300 -
JO
1
400 -
• " i
1
10
Wavelength (m)
Figure 17. A. Power spectral density plot of logging velocity values for the
ooze interval (80-350 mbsf) for Site 806. The plot shows dominant power at
a wavelength of 9.5 m and its harmonics (4.75 and 3.17 m). This plot is used
as an example to demonstrate that most of the variation in logging velocity data
through the ooze interval is probably a result of coring disturbance (widening
of the borehole during the coring process), because the advanced hydraulic
piston core (APC) is precisely 9.5 m in length. This effect can be seen in the
logging velocity data in the ooze intervals at each of the Leg 130 drill sites. B.
A power spectral density plot of logging gamma-ray density values at Site 806,
which also demonstrates the 9.5-m wavelength power resulting from coring
disturbance. In this case, however, the harmonic wavelengths are not present,
either because they are probably dampened by real data or because the log
density values are partly corrected for borehole diameter.
500 -
600 Log velocity
700 -
Corrected lab velocity
800
Figure 16. Plot of final corrected laboratory P-wave velocity and logging
/>-wave velocity vs. depth at Site 806, showing that lab values are now
congruent with log values. The lab values are substituted for log data over the
ooze interval.
With velocity and density data available, it is possible to generate
synthetic seismograms to assist in the correlation of core data to the
seismic data. Correlation of the synthetic seismograms to the field
seismic records suggests that the true in-situ velocities are 30 m/s
lower than those measured by sonic logging velocity analysis. These
correlations permit the conversion of seismic traveltimes to depths
below seafloor at each site (Table 1). Plotting this data, it was found
that a single velocity-depth function exists for all sites on the Ontong
Java Plateau, allowing a single equation to represent the transfer of
traveltime to depth or vice versa, anywhere on the plateau (Fig. 18).
This fact implies that depth of burial is the primary control on
sediment compressional wave velocity, rather than age, or state of
induration, which can vary from site to site. This finding agrees with
the methods used in the correction of velocity to in-situ conditions,
where velocity was corrected as a function of depth below seafloor,
rather than as a function of any of the physical properties. It also
supports the implication that the elastic moduli (primarily rigidity)
45
D.C. MOSHER ET AL.
800
700
600
•
Sitθ586
α
Site806
o
Site 807
×
Site805
+
Site803
α×
Q 300
ox"
200
100
+ o×
i
100
D
Depth (mbsf) = 44.49 + 0.800x + 3.308X10-4x^
Traveltime (ms) = -35.185 +1.118y -1.969X10' 4 y 2
200
300
400
500
i
600
700
Depth on seismic profile (ms two-way traveltime)
Figure 18. Cross plot of seismic stratigraphy and synthetic seismogram correlation results for each of the drill sites. Depth in meters is the ordinate axis, and
traveltime below seafloor in milliseconds is the abscissa. Note that results from each of the sites plot on a single curve, suggesting that the velocity-depth function
is nearly equivalent at each site. A second-order polynomial fit provides for converting traveltime on the seismic section (measured below the first high-amplitude
seafloor return, not on the precursor) to depth in meters below seafloor throughout the study area. The equation of the fit is given, as is the inverse equation for
converting depth to traveltime.
are the primary physical controls on the velocity of Ontong Java
Plateau sediment. Because the sediments are of one type (carbonate),
little lithologic control is exerted on velocity; and, under pressure
from depth of burial, it seems that the sediments behave similarly,
regardless of state of induration.
Seismic reflection profiles from the Ontong Java Plateau show
three distinct changes in character with increasing water depth: (1) excessive thinning of the sediment column, (2) changes in the echocharacter (seismic facies), and (3) loss of continuous, coherent
reflections. From the top of the Ontong Java Plateau down to about
2900 m water depth, the sediment column is about 1 s thick and is
characterized by numerous, parallel, continuous, coherent, low-amplitude reflections in seismic profile. Correlation with drilling results
indicates that these sediments were deposited by pelagic processes.
The sediment is composed of about 90% biogenic carbonate material.
Between 2900 and 3375 m water depth, seismic profiles become
more complex, with a variety of reflection patterns. Faulting, reflector pinchout and truncation, and totally incoherent reflections characterize this section of the flank. The sediment column thins
significantly (25%) through this interval. Seismic and core evidence
indicate that these sediments are composed of pelagic carbonate;
however, in places they are moderately to heavily modified by
mass-movement processes.
Between 3375 and 3800 m water depth, seismic reflectors become
more coherent and of higher amplitude, but faulting and resedimentation are still apparent on seismic profiles. Core evidence indicates
that sedimentation was dominated by pelagic deposition, but sediment
reworking and dissolution have occurred. High-amplitude reflections
and variability in the seismic profiles support these findings.
Below 3800 m, acoustic basement is characterized by a rough
topography, possibly related to basement faulting, creating a series of
horsts and grabens (Hagen et al., this volume). Sediments tend to infill
the grabens. The sediment column yields reflections of different
characteristics, lending to development of several seismic facies.
Coring in one of these basins yielded a stratigraphy with a number of
hiatuses, debris flows, and disturbed sediments. These interpretations
support inferences made from the seismic records. It seems that much
46
of the sedimentary record is missing or condensed. Physical property
contrasts, as a result, are high, leading to high-amplitude reflections
in the seismic profile.
The flank of the Ontong Java Plateau levels off in the Nauru Basin
at about 4500 m water depth. Acoustic basement is smoother at this
point. The sediment column is about 25% as thick as on top of the
plateau. No sampling has been conducted at this depth, but seismic
sections provide evidence of turbidite deposition and seafloor erosion.
Seismic stratigraphic analysis has shown that the sediment section
decreases in thickness by 65% from 2200 m water depth on top of the
plateau to 4000 m water depth near the base of the plateau. Thinning
is attributed to increasing carbonate dissolution with depth, which is
a function of the lysocline position and the CCD. If this assumption
is true, then changes in the relative thicknesses of seismostratigraphic
units at each drill site are indicative of changes in the position of the
lysocline and the dissolution gradient between the lysocline and the
CCD. We think it possible that a shallow lysocline in the early
Miocene caused sediment thinning. Deepening of the lysocline in the
late-early Miocene caused relative thickening at each site. Within the
middle Miocene, a sharp rise occurs in the lysoclinal depth, concurrent with a possible steepening of the dissolution gradient, which
resulted in sediment thinning at all four sites. The thicker sections in
the late Miocene likely correspond to a deepening of the lysocline,
and a subsequent rise in the lysocline again hinders accumulation of
sediment in the very late Miocene and Pliocene.
ACKNOWLEDGMENTS
The authors would like to expressly thank the officers, crew, and
scientific staff of JOIDES Resolution during Leg 130 and of Thomas
Washington during Roundabout Leg 11. We would also like to thank
Stephan Saustrup, Brian Nichols, and Peter Bugden for technical
support. The Canadian Secretariat for the Ocean Drilling Program
supplied travel funding for D. Mosher. D. Mosher was supported with
a NSERC postgraduate scholarship and Dalhousie University fellowship. Support for this work was supplied in part by the Office of Naval
Research Contract No. n00014-91-1213 and seismic data collection
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
Table 1. Depths to reflectors on thefieldseismic data at each
of the site crossings, correlations with synthetics at each site,
and age assignments.
Traveltime (ms)
Depth
(mbsf)
Age
(Ma)
148
225
310
360
380
116
181
257
307
327
6.8
9.2
17.3
24.5
25.7
105
185
260
388
455
525
165
250
328
439
520
590
128
198
266
378
467
548
5.05
7.18
9.1
16.3
23.7
26.4
Rl
R2
R3
R4
R5
R6
120
225
325
515
600
675
195
295
400
592
690
765
156
244
343
546
665
754
4.8
6.8
9.1
16.3
22.6
25.2
807
Rl
R2
R3
R4
R5
R6
92
190
280
410
490
560
154.5
250.9
338.6
460
554
616.6
125
210
293
416
521
592
5.0
6.9
9.5
16.1
22.2
24.5
289/586
Rl
R2
R3
R4
R5
R6
130
225
315
490
575
655
195
290
375
555
640
725
160
246
328
519
620
713
5.0
7.2
9.7
17.0
22.7
Site
Reflector
Seismic
Rl
R2
R3
R4
R5
R6
75
165
250
300
315
805
Rl
R2
R3
R4
R5
R6
806
803
Synthetic
Note: Age assignments come from Leg 130 biostratigraphic data
(Kroenke, Berger, Janecek, et al., 1991).
was funded by a NSF-ODP grant to E.L. Winterer. The digital sound
velocimeter (DSV) for shipboard velocity measurements was provided by the Département de Géologie Dynamique, Université Pierre
et Marie Curie, and funded to Y. Lancelot by INSU/CNRS, France.
REFERENCES*
Berger, W.H., 1967. Foraminiferal ooze: solution at depths. Science, 156:383-385
, 1968. Planktonic foraminifera: selective solution and paleoclimatic
interpretation. Deep-Sea Res., Pt. A, 15:31-45.
-, 1970. Planktonic Foraminifera: selective solution and the lysocline.
Mar. Geol, 8:111-138.
-, 1989. Appendix: global maps of ocean productivity. In Berger,
W.H., Smetacek, V.S., and Wefer, G. (Eds.), Productivity of the Ocean:
Present and Past: New York (Wiley), 429-455.
Berger, W.H., Fischer, K., Lai, C, and Wu, G., 1988. Ocean carbon flux: global
maps of primary production and export production. In Agegian, C.R. (Ed.),
Biogeochemical Cycling and Fluxes Between the Deep Euphotic Zone and
Other Oceanic Realms. NOAANational Res. Prog., Res. Rep. 88-1:131-176.
Berger, W.H., and Johnson, T.C., 1976. Deep sea carbonates: dissolution and
mass wasting on Ontong-Java Plateau. Science, 192:785-787.
Berger, W.H., Johnson, T.C., and Hamilton, E.L., 1977. Sedimentation on
Ontong-Java Plateau: observations on a classic "carbonate monitor." In
Anderson, N.R., and Malahoff, A. (Eds.), The Fate of Fossil Fuel CO2 in
the Oceans: New York (Plenum), 543-567.
* Abbreviations for names of organizations and publication titles in ODP
reference lists follow the style given in Chemical Abstracts Service Source
Index (published by American Chemical Society).
Berger, W.H., and Mayer, L. A., 1978. Deep-sea carbonates: acoustic reflectors
and lysocline fluctuations. Geology, 6:11-15.
Boyce, R.E., 1973a. Physical Properties—methods. In Edgar, N.T., Saunders,
J.B., et al., Init. Repts. DSDP, 15: Washington (U.S. Govt. Printing Office),
1115-1128.
, 1973b. Summary of physical properties. In Edgar, N.T., Saunders,
J.B., et al., Init. Repts. DSDP, 15: Washington (U.S. Govt. Printing Office),
1067-1068.
-, 1976. Sound velocity-density parameters of sediment and rock
from DSDP drill Sites 315-318 on the Line Islands Chain, Manihiki
Plateau, and Tuamotu Ridge in the Pacific Ocean. In Schlanger, S.O.,
Jackson, E.D., et al., Init. Repts. DSDP, 33: Washington (U.S. Govt.
Printing Office), 695-728.
Carlson, R.L., Christensen, N.I., and Moore, R.P., 1980. Anomalous crustal
structures in ocean basins: continental fragments and oceanic plateaus.
Earth Planet. Sci. Lett., 51:71-180.
Courtney, R., and Mayer, L.A., in press. Acoustic properties of fine-grained
sediments from Emerald Basin: towards an inversion of physical properties
using Biofs law. J. Acoustic. Soc. Am.
Das, D.M., 1983. Advanced Soil Mechanics: Washington (Hemisphere Publ.).
Farrell, J.W., and Prell, W.L., 1989. Climatic change and CaCO3 preservation:
an 800,000 year bathymetric reconstruction from the central equatorial
Pacific Ocean. Paleoceanography, 4:447-466.
Fulthorpe, C.S., Schlanger, S.O., and Jarrard, R.D., 1989. In-situ acoustic
properties of pelagic carbonate sediments on the Ontong Java Plateau. J.
Geophys. Res., 94:4025-4032.
Furumoto, A.S., Webb, J.P., Odegard, M.E., and Hussong, D.M., 1976. Seismic
studies on the Ontong Java Plateau, 1970. Tectonophysics, 34:71-90.
Hamilton, E.L., 1970. Reflection coefficients and bottom losses at normal
incidence computed from Pacific sediment properties. Geophysics, 35:995.
, 1971a. Elastic properties of marine sediments. J. Geophys. Res.,
76:579-604.
-, 1971b. Prediction of in-situ acoustic and elastic properties of
marine sediments. Geophysics, 36:266-284.
-, 1972. Compressional wave attenuation in marine sediments. Geophysics, 37:620-646.
-, 1974. Geoacoustic models of the sea floor. In Hampton, L. (Ed.),
The Physics of Sound in Marine Sediments: New York (Plenum),
181-221.
-, 1980. Geoacoustic modelling of the sea floor. J. Acoust. Soc. Am.,
68:1313-1340.
Hamilton, E.L., Bachman, R.T., Berger, W.H., Johnson, T.C., and Mayer, L.A.,
1982. Acoustic and related properties of calcareous deep-sea sediments. J.
Sediment. Petrol, 52:733-753.
Houtz, R.E., and Ludwig, W.J., 1979. Distribution of reverberant subbottom
layers in the southwest Pacific basin. /. Geophys. Res., 84:3497-3304.
Hussong, D.M., Wipperman, L.K., and Kroenke, L.W., 1979. The crustal
structure of the Ontong Java and Manihiki Oceanic Plateaus. J. Geophys.
Res., 84:6003-6010.
Johnson, D.A., and Johnson, T.C., 1970. Sediment redistribution by bottom
currents in the central Pacific. Deep-Sea Res., Pt. A, 17:157-169.
Kroenke, L.W, 1972. Geology of the Ontong-Java Plateau [Ph.D. dissert.].
Univ. of Hawaii, Honolulu.
Kroenke, L.W, Berger, WH, Janecek, T.R., et al., 1991. Proc. ODP, Init.
Repts., 130: College Station, TX (Ocean Drilling Program).
Kroenke, L.W, Resig, J.M., and Cooper, P.A., 1986. Tectonics of the southeastern Solomon Islands: formation of the Malaita Anticlinorium. In
Vedder, J.G., Pound, K.S., and Boundy, S.Q. (Eds.), Geology and Offshore
Resources of Pacific Island Arcs, Central and Western Solomon Islands.
Circum-Pac. Counc. Energy Mineral Resour., Earth Sci. Ser, 4:109-116.
Langseth, M.G., Nalone, I., and Berger, D., 1971. Seafloor geothermal measurements from Vema Cruise 24. Columbia Univ. Tech. Rept., 3-CU-3-71.
Laughton, A.S., 1957. Sound propagation in compacted ocean sediments.
Geophysics, 22:233-260.
Mayer, L.A., Courtney, R.C., and Moran, K., 1987. Ultrasonic measurement
of marine sediment properties. Proc. Oceanogr., 87:1-139.
Mayer, L.A., Shipley, T.H., Theyer, F., Wilkens, R.H., and Winterer, E.L.,
1985. Seismic modeling and paleoceanography at Deep Sea Drilling
Project Site 574. In Mayer, L., Theyer, E, Thomas, E., et al., Init. Repts.
DSDP, 85: Washington (U.S. Govt. Printing Office), 947-970.
Mayer, L.A., Shipley, T.H., and Winterer, E.L., 1986. Equatorial seismic reflectors
as indicators of global oceanographic events. Science, 233:761-764.
Mayer, L.A., Shipley, T.H., Winterer, E.L., Mosher, D.C., and Hagen, R.A.,
1991. Sea Beam and seismic reflection surveys on the Ontong Java Plateau.
47
D.C. MOSHER ET AL.
In Kroenke, L.W., Berger, W.H., Janecek, T.R., et al., Proc. ODP, Init.
Repts., 130: College Station, TX (Ocean Drilling Program), 45-75.
Moore, T.C., and Heath, G.R., 1967. Abyssal Hills in the central equatorial
Pacific: detailed structure of the seafloor and subbottom reflectors. Mar.
Geol, 5:161-179.
Mudie, J.D., Grow, J.A., and Bessey, J.S., 1972. Anear-bottom survey of lineated
abyssal hills in the equatorial Pacific. Mar. Geophys. Res., 1:397—411.
Resig, J.M., Buyannanonth, V., and Roy, K.J., 1976. Foraminiferal stratigraphy
and depositional history in the area of the Ontong-Java Plateau. Deep-Sea
Res., Pt. A, 23:441-456.
Shipboard Scientific Party, 1971. Site 64. In Winterer, E.L., Riedel, W.R., et al.,
Init. Repts. DSDP, 7, Pt. 1: Washington (U.S. Govt. Printing Office), 473-606.
, 1975. Site 289. In Andrews, J.E., Packham, G., et al., Init. Repts.
DSDP, 30: Washington (U.S. Govt. Printing Office), 231-398.
Shipboard Scientific Parties of Legs 89 and 90,1986. Site 586: western equatorial
Pacific. In Kennett, J.P., von der Borch, C.C., et al., Init. Repts. DSDP, 90:
Washington (U.S. Govt. Printing Office), 19-114.
Shipley, T.H., Mayer, L.A., and Winterer, E.L., 1989. Seismic reflections and
sedimentation in the pelagic equatorial Pacific. In Bally, A.W. (Ed.), Atlas
of Seismic Stratigraphy. AAPG, 3:224-231.
Shipley, T.H., Winterer, E.L., Goud, M., Hills, S J., Metzler, C.V., Paull, C.K., and
Shay, J.T., 1985. Sea Beam bathymetric and water-gun seismic reflection
surveys in the equatorial Pacific. In Mayer, L., Theyer, E, Thomas, E., et al.,
Init. Repts. DSDP, 85: Washington (U.S. Govt. Printing Office), 825-837.
van Andel, T.H., Heath, G.R., and Moore, T.C., Jr., 1975. Cenozoic History
and Paleoceanography of the Central Equatorial Pacific Ocean. Mem.
Geol. Soc. Am., 143.
Wilson, W.D., 1960. Speed of sound in sea water as a function of temperature,
pressure and salinity. J. Acoustic. Soc. Am., 32:641-644.
Wood, A.B., 1941. A Textbook of Sound: New York (MacMillan Press).
18
Wu, G., and Berger, W.H., 1991. Pleistocene O records from the Ontong-Java
Plateau: effects of winnowing and dissolution. Mar. Geol, 96:193-209.
Wu, G., Yasuda, M.K., and Berger, W.H., 1991. Late Pleistocene carbonate
stratigraphy on Ontong Java Plateau in the western equatorial Pacific. Mar.
Geol, 99:135-150.
Wyllie, M.R.J., Gregory, H.R., and Gardner, L.W., 1956. Elastic waves in
heterogeneous and porous media. Geophysics, 22:41.
Date of initial receipt: 2 December 1991
Date of acceptance: 10 August 1992
Ms 130B-047
B
800 -
Correlati i/e depths ( •nbst)
7OO 600 -
800
Lyle θt al. (this volume)
correlations
• Seismic
correlations
7OO -
Site 807
C 600XI
500 4OO -
Lyle et al. (this volume)
correlations
Seismic
correlations
Site 807
^T 500
I
Site 586
Site 806
"O 4OO —
I
300 -
30
-§ ° ~
Site 803
200 -
0
Site 803
200 -
100 -
100 ~
I
200
I
400
I
600
Depth (mbsf) at Site 806
800
200
400
I
600
800
Depth (mbsf) at Site 586
Figure 19. A. Comparison plot of correlative depths at Site 806 vs. other sites, using seismic stratigraphy correlations (traveltimes converted to depths below
seafloor using synthetic correlations), and Lyle et al. (this volume) correlations using an inverse signal correlation technique, comparing sites based on the
shape of logging density and resistivity curves. B. Equivalent to Figure 19A, but depths at Site 586 are used as abscissa values. The correlations of Lyle et
al. (this volume) are given in greater detail (about 1-m spacing) than the seismic correlations, which are based on a maximum of six horizons, spaced
throughout the sediment column.
SEISMIC STRATIGRAPHY—ONTONG JAVA PLATEAU
140
120
-
100
-?
80
θO
-
40
-
20
-
R3
R6
R4
Reflection horizon
Bsmt
Figure 20. Plot of percent thickness of seismostratigraphic units (i.e., seafloor to Rl, Rl to R2, R2 to R3, R3 to R4, R4 to R5, R5 to R6, and R6 to
acoustic basement [Bsmt]) at each drill site compared with thicknesses at Site 586. If a value is >100%, it is thicker than at Site 586; if <100%, it
is thinner. Thickness measurements were derived from depth in meters below seafloor, which in turn were derived from the results of the
synthetic-seismic correlations (see Table 1). The curves are cubic spline fits through the data points. We think these thickness measurements can
be used to map out the relative positions of past lysoclines.
Table 2. Summary of ages assigned to seismic
reflectors in Table 1.
3500
2200
Lysoclinβ (this study)
-3900
2600-
-4300
Reflector
Age
(Ma)
Rl
R2
R3
R4
R5
R6
4.8-5.05
6.8-7.2
9.1-9.7
16.1-17.3
22.2-24.5
24.5-26.4
Epoch
Pliocene/Miocene boundary
late/middle Miocene boundary
middle/early Miocene boundary
Miocene/Oligocene boundary
4700 Q
8
3800-
4200
CCD for central Pacific
(after van Andel et al., 1975)"
1
i
5
10
-5100
i
i
i
15
20
25
5500
30
Age (Ma)
Figure 21. Approximate position of the lysocline through the Neogene as
determined by the amount of relative thinning apparent from seismic stratigraphy data through the drill sites. To produce this curve, we have assumed that
sediment thinning is caused entirely by carbonate dissolution, and that water
depths of the drill sites are about the same through the Neogene as at present.
The CCD depths, as reported by van Andel et al. (1975) for the central
equatorial Pacific, are plotted with this curve. The two curves bear little
resemblance to one another.
49